Brazil Investment Report 2009 — The Credibility Dividend

Brazil's 2009 is a case study in what institutional credibility is worth, and the value is easiest to see by comparison with its own past.

In previous global shocks — through the 1980s and 1990s — a crisis in developed markets produced a specific and painful sequence in Brazil: capital left, the currency collapsed, inflation surged, and the policy response was forced to be contractionary — raising rates and cutting spending into a downturn, in order to defend the currency and contain inflation. The economy amplified the external shock rather than absorbing it.

In 2009 the sequence was different. Capital left and the currency fell, but inflation expectations stayed anchored, which meant the central bank could cut rates rather than raise them. The government could increase spending rather than cut it. For the first time, the policy response worked against the shock instead of with it.

This capacity was not a decision made in 2009. It was the accumulated return on institutional work done over the preceding fifteen years:

• An inflation-targeting framework with a track record, so a currency fall was not automatically assumed to become inflation.
• A floating exchange rate, so the currency could absorb the shock rather than being defended until reserves ran out.
• Substantial foreign exchange reserves, accumulated during the preceding boom — the insurance the Asia-Pacific Investment Report 2008 describes, in a different hemisphere.
• Reduced foreign currency government debt, so a currency fall no longer multiplied the debt burden.
• And a sound, well-capitalised, well-supervised banking system carrying none of the structured credit exposure at the centre of the crisis.

Each of these had a visible cost during the good years and no visible benefit. 2009 is when the benefit arrived, all at once.

The vulnerabilities the year created are the report's second half: a growing dependency on one economy's commodity demand, a currency appreciating to the detriment of manufacturing, and a large expansion of state-directed credit that proved much easier to start than to stop.

What policy space actually is

The phrase is used loosely. It has specific components, all of which are built in advance and none of which can be acquired during a crisis.

Monetary space requires that inflation expectations be anchored. A central bank can only cut rates in a currency crisis if the public believes it will not permit inflation. Without that belief, a rate cut confirms fears and accelerates the currency's fall — so the instrument is unavailable precisely when it is needed. The belief takes years of demonstrated behaviour to establish and can be lost quickly.

Fiscal space requires a debt level and structure that permit borrowing more without triggering doubt about repayment. This is a function of the debt ratio, its maturity, its currency composition, and who holds it. Short-dated foreign currency debt held by foreigners provides almost no space; long-dated domestic currency debt held domestically provides a great deal.

External space requires reserves sufficient to meet obligations without forced adjustment, per the metrics in the Global Investment Outlook 2013.

Financial space requires a banking system that can keep lending. A system needing recapitalisation transmits the shock rather than absorbing it, which is what distinguished the developed economies in this crisis.

Every component of policy space is accumulated during good times, at a visible cost, for a benefit that only appears in a crisis. This is the same asymmetry as insurance, and it produces the same systematic under-provision.

Why Brazil had built it is worth stating plainly: it had experienced the alternative repeatedly and recently. The institutional reforms of the 1990s were responses to hyperinflation and to currency crises whose costs were within living memory of everyone making the decisions. The lesson had been paid for.

This is the archive's most uncomfortable finding about learning: economies build resilience in proportion to how recently they were damaged, and the memory decays. The Asia-Pacific Investment Report 2008 documents the same mechanism — post-1997 reserve accumulation that looked wasteful until it did not.

Why an external shock with sound balance sheets recovers fast

The recovery's speed is explicable from the structure of the shock, and the framework is useful for any similar situation.

The developed economies faced a balance sheet recession, per the Global Investment Outlook 2010: households, banks and firms all needed to repair, and repair takes years regardless of policy.

Brazil faced something different: an external demand shock hitting sound balance sheets.

The distinction matters enormously:

Balance sheet recession	External demand shock
Domestic agents over-indebted	Domestic agents sound
Repair required before growth	No repair required
Policy transmission impaired	Policy transmission works
Recovery measured in years	Recovery measured in quarters

Why recovery is fast when balance sheets are sound:

• Nobody needs to deleverage, so lower rates translate into borrowing rather than into debt repayment.
• Banks can lend, because their capital is intact.
• And the shock itself is temporary — external demand returns when the exporting economies recover, and in this case a large stimulus elsewhere accelerated that.

The commodity channel amplified the recovery in exactly the way the Asia-Pacific Investment Report 2009 describes from the other side. The stimulus programme in Asia drove enormous demand for iron ore, soybeans, oil and other exports, and prices recovered far faster than developed-world demand did.

So Brazil's recovery had two engines: functioning domestic policy transmission, and a commodity price recovery driven by another economy's construction programme.

The second engine is where the vulnerability was being created, and almost nobody was describing it as a dependency at the time.

Replacing one dependency with another

The structural shift of this period deserves clear statement because it was widely celebrated and rarely analysed as a concentration.

The old dependency was on developed-world demand and on developed-world capital — the channels that had transmitted every previous crisis.

The new dependency was on a single economy's demand for a narrow set of commodities.

Why this is more concentrated, not less:

• The customer base narrowed. Developed-world demand is spread across many economies with imperfectly correlated cycles. A single large buyer is one decision-maker.
• The product base narrowed. Commodity export revenue is concentrated in a few products whose prices move together, since they respond to the same industrial demand.
• And the demand driver was a policy programme, not a structural trend — the attribution error the Asia-Pacific Investment Report 2009 identifies.

The reason it was read as diversification is that the geographic label changed. Trade shifted from one region to another, which looks like diversification on a map and is the opposite in exposure terms.

Replacing many customers buying many things with one customer buying few things is a concentration, whatever it does to the geographic distribution of the trade statistics.

The investment consequence was that Brazilian assets became a leveraged position on another economy's investment cycle — equity, currency and fiscal position all responding to the same driver.

And the fiscal position was the most exposed of the three, because commodity revenue flowed into the budget and spending expanded to use it. This is the fiscal breakeven problem the Asia-Pacific Investment Report 2014 sets out, and it determined how the eventual downturn was experienced.

The currency squeeze on manufacturing

A slower-moving problem was developing throughout, and its mechanism is worth distinguishing from the standard account.

The classic resource-driven adjustment runs through the trade account: a commodity boom raises export earnings, the currency appreciates, and other tradeable sectors lose competitiveness.

In this period the financial account did as much work as the trade account:

• High domestic interest rates, maintained for inflation-targeting reasons, attracted carry trade capital — investors borrowing cheaply in developed markets, per the Global Investment Outlook 2010, and investing at high domestic rates.
• This inflow appreciated the currency independently of the trade position.
• And the appreciation was self-reinforcing for a period, since currency gains added to the carry trade's return, attracting more capital.

The consequence for manufacturing:

• Exports became uncompetitive against producers in economies with weaker currencies.
• Imports became cheap, taking domestic market share.
• And the effect compounded with high domestic borrowing costs, so manufacturers faced expensive capital and an unfavourable exchange rate simultaneously.

The policy bind is the one the Asia-Pacific Investment Report 2010 describes, arriving here in a more acute form: rates could not be cut to weaken the currency without abandoning the inflation credibility that made the whole policy framework work. The credibility that had provided the crisis response was constraining the recovery's composition.

The responses attempted — taxes on financial inflows, intervention, and later a deliberate rate-cutting cycle — had mixed results, and the manufacturing share of output continued declining.

This is a genuine trade-off rather than a policy error. Inflation credibility is extremely valuable, as the crisis response demonstrated. Its cost was a structurally strong currency and a smaller manufacturing base, and no available instrument delivered both.

Directed credit and the exit problem

The credit response deserves attention because its design created a lasting fiscal position.

The mechanism: with private credit contracting, state-controlled banks expanded lending sharply, funded in part by government transfers. This substituted public credit for private credit and kept investment going.

As crisis policy this worked. It was fast, large and directed at exactly the constraint — the mechanism the Asia-Pacific Investment Report 2009 identifies as investment being chosen for controllability.

The problems were structural and appeared later:

• The lending was frequently at subsidised rates, below the government's own cost of borrowing. The difference is a continuing fiscal cost, and it compounds for the life of each loan.
• The subsidy weakened monetary policy. A large share of credit priced below the policy rate does not respond to changes in the policy rate, so the central bank must move rates further to achieve the same effect on the responsive share.
• The allocation was directed, so capital flowed toward favoured sectors and large firms rather than by return — the allocation problem the Asia-Pacific Investment Report 2012 describes in a different context.
• And it was very difficult to withdraw. Borrowers had built plans around subsidised credit, and removing it meant a visible tightening for identifiable firms. Concentrated losers, diffuse gainers.

A crisis credit programme is easy to start because the emergency justifies it and easy to continue because the beneficiaries are organised. The exit problem is the same one forward guidance creates, in a different instrument.

The Global Investment Outlook 2013 describes the identical exit bind in monetary policy — a commitment that works because it is credible becomes costly to withdraw precisely because people acted on it.

What an allocator could act on

Assess policy space by its components, built in advance. Anchored inflation expectations, debt structure, reserve adequacy and banking system capital are all measurable and none can be acquired during a crisis.

Classify the shock before forecasting the recovery. An external demand shock hitting sound balance sheets recovers in quarters; a balance sheet recession takes years, and the same policy produces different results in each.

Read geographic diversification of trade as a hypothesis, not a fact. Shifting from many customers buying many products to one customer buying few is a concentration however it appears on a map.

Check whether commodity demand is policy-driven or structurally driven. They are indistinguishable in price data and imply completely different durations.

Look at the fiscal breakeven, not the fiscal balance, for commodity exporters. A budget balanced at boom prices is a structural deficit concealed by the cycle.

Measure the share of credit priced below the policy rate. Directed lending at subsidised rates does not respond to monetary policy, so the central bank must move rates further to reach the responsive remainder.

What 2009 established

• Counter-cyclical capacity is an accumulated asset, and Brazil could respond in 2009 only because of reforms made years earlier at visible cost.
• External shocks to sound balance sheets recover fast, in contrast to the balance sheet recessions elsewhere.
• The commodity relationship replaced a diversified dependency with a concentrated one, misread as diversification because the geography changed.
• Inflation credibility carried a currency cost that fell on manufacturing, with no instrument able to deliver both.
• Directed credit worked as crisis policy and created a durable fiscal exposure with the same exit problem as forward guidance.